A Study in Babylonian and Assybian Archaeology

XVI

GESTURE IN SUMERIAN AND BABYLONIAN PRAYER

A STUDY IN BABYLONIAN AND ASSYBIAN ARCHAEOLOGY

BY S. LANGDON, PH.D., PROFESSOR OF ASSYRIOLOGY, OXFORD ELIGIOUS worship is abundantly illustrated in R many of its most important aspects by scenes engraved on Sumerian, Babylonian, and Assyrian seal cylinders. Chronologically the seals of this region illustrate nearly every period of the long history of these peoples and the changing rituals and beliefs of their religion. A very large proportion of the seals represent the owner of the seal approaching a deity in the attitude of prayer. This is especially true of the glyptique of Sumer and Akkad, where the proportion of this type of seal to all others is much greater than in Assyria. In the northern empire the Assyrians are not so much attached to the scenes of worship, but even here this motif is well represented. The engravers of cylinders in all periods probably kept in stock seals engraved with the scene of the private prayer as the custom imposed in their periods. The human who is figured standing before a god, or in Assyria more frequently before a divine symbol, is not a portrait of the owner of the seal. The owner regards himself rather as represented and symbolized by the conventional figure. In those cases in which the engraver produced a seal cylinder at the command of a Sumerian or Babylonian, perhaps, we may regard the praying figure as an approximate portrait. That scenes of this kind are standardized products of the various

Downloaded from https://www.cambridge.org/core. University of Warwick, on 20 May 2018 at 01:28:38, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/S0035869X00053296 532 GESTURE IN SUMERIAN AND BABYLONIAN PRAYER

periods representing the religious ideas, but not actual portraits, is proven by Cassite seals of women on which a male figure takes the place of the worshipper.1 The example of a portrait of a woman worshipper on Ward No. 536 would in itself prove that the praying figures on Babylonian seals actually represent the owners. But even more direct evidence for this important fact may be adduced. The Aramaic traders and adventurers who settled in Assyria and Babylonia in great numbers in the late period adopted the cylinder seal. They obviously purchased these from Assyrian and Babylonian engravers, whose designs are purely of the accepted type. The owners probably cared nothing for the religious symbolism and scenes on the seals. They worshipped other gods and adhered to other forms of religion. But in many cases these Aramaic citizens of Assyria caused their names to be inscribed in Aramaic letters beside the design. On the seal reproduced here (Fig. 1) from about the seventh century B.C. we have typical Assyrian symbols. Winged genii adore the winged disk of the sun-god. The owner of the seal stands supplicating the forked lightning, symbol of Adad, the thunder-god. Beside the design is written in Aramaic, "Jarp'el, son of Hur-'adad." To ensure the identity of the figured person the engraver repeats the name (which is engraved very closely beside the figure) " Jarp'el ". This evidence adds a very important fact to our knowledge of Babylonian religion. The praying figures on seals actually represent the owners. Of that we can no longer doubt. Sumerians, Babylonians, and Assyrians carried about on their seals representations of themselves as they said their prayers before one of the great gods.

1 Ward, Seal Cylinders of Western Asia, No. 536, is an example of a1 portrait. The seal belongs to the woman Meuarubtum and the praying figure is a woman. But see Collection de Clercq, 262, seal of Ussurtum with male figure in the scene. See also ibid. 265, seal of Tdh-nis-resi.

These were supported from the neck by a stout cord which passed through an aperture at the axis of the cylinder. I propose here to study the various attitudes of the worshipper's hands in the different periods, and to compare these attitudes with those which characterize the worship of adjacent peoples. The early period of Sumerian glyptique, commonly known as pre-Sargonic, lias generally the so-called processional scene. This will be illustrated by Figs. 2, 3, 4, and 5. Seals of this type represent the owner conducted into the presence of a great seated deity by his own personal god, who leads his protege by the hand. In case the procession moves from left to right the man's god or goddess takes the worshipper's left hand. In the reversed direction the man is led by the right hand. By this design the artist brings the disengaged arm nearest the observer. Occasionally the disengaged arm is employed to carry a lamb or kid as an offering. On Fig. 3 the owner is conducted by his goddess. On Fig. 2 an attendant brings the animal sacrifice; the reader will observe that this attendant approaches with the right arm extended and the fore-arm raised parallel with the face, palm inward. Observe also that the conducting deities approach with disengaged arm raised in a similar manner palms inward. On seal Fig. 7 three deities approach the seated grain goddess. The central figure (a goddess) of these three has the most ancient attitude of prayer for humans, the raised hand palm inward and the disengaged arm folded at the waist. These are all extremely archaic types extending back to a period as early as 3500 B.C. From them we conclude that man when not conducted by a deity stood in the position of prayer described above. This is apparently the original prayer attitude of prehistoric man in Sumer. For this attitude of primitive man see also the following extremely archaic seals, Delaporte, BMiotheque Nationale, 51, 53, 59;

Ward, 90, 91, 302; Delaporte, Musee Guimet, 25, 26; Collection de Clercq, 83.1 Although such was the orthodox pose of adoration in the most ancient times a few exceptions from the same period should be noted. Fig. 6, a product of the engraver's art well toward the end of the pre-Sargonic period, say 2900 B.C., shows the human with both hands folded at the waist. His god, double-faced, and goddess precede him in the ordinary pose of the period. On a much more ancient seal, Musee Guimet, 23, the owner stands in the same attitude, both hands at the waist. A seal to be placed at the end of the archaic period has two humans, a bearded Semite owner of the seal attended by a Sumerian priest (?). The owner is in the second pose, and so is an inferior deity who stands behind the seated goddess (drawn double vis-a-vis). The priest has the orthodox position with the modification that the palm is not turned inward but faces the left (De Clercq, No. 82). This later modification of the ordinary pose will be seen on Fig.'8. The hand is thus brought into such position that the narrow surface on the side of the little finger is turned toward the deitjr. It is possible that the seals Figs. 7 and 8 may be assigned to the period of Sargon and Narain-Sin. The second or Sargonic period of glyptique, say 2800—2600, does not introduce the scene of adoration and prayer to any great extent. Scenes from the GilgamishEpic are by far the most common here. On a seal of Naram- Sin, Rev. d'Assyr., iv, 11, the praying figure has the older orthodox position. To this period belongs the scene of a row of inferior deities adoring the seated sun-god (Delaporte, Bibliotheque Nationale, 72, 63, 64). The deity who heads the procession has nearly the old orthodox pose, the others have the secondary position with both

1 Seal of Gimil-l-U-su , . . li Mz-luh-ka-ki, "(jimililisu tlie ... of Meluhha."

hands folded at the waist. Both attitudes, therefore, obtain in the religious rituals of early Sumerian and Semitic civilization. For the Semitic pose in prayer the cylinders prove nothing, for this people simply adopted the Sumerian custom. It will be seen that the custom of raising the hand palm inward was apparently the most ancient and universal, that is, the kiss-throwing hand. Judging from those scenes in which both poses appear, one is induced to believe that the hands folded at the waist indicated an attitude of great humility and penance. The lifted hand, on the contrary, would lay the emphasis on adoration and salutation. The third period of glyptique includes the seals of the (•schools of Gudea and Dungi, roughly 2600-2358, or down to the end of the dynasty of Ur.1 If we may make inferences from the two seals of Gudea (Figs. 9, 10), only the processional scene was admitted on seals in the period following upon the Sargonic dynasty. This is, of course, a pure fancy, and does not enable us to see the real attitude of prayer adopted at this stage of Sumerian religion. Here the worshipper lifts his disengaged arm in exactly the same position as that adopted in the oldest pose. The inference is that in actual worship the Sumerians of the Gudea period stood with right forearm raised before the face palm inward. On Fig. 9 the interceding figure of the patesi's goddess has both hands raised, palms at right angle to the face. This position becomes, henceforth, orthodox for the ubiquitous figure of the weeping mother goddess on seals of the Ur, Isin, and Babylonian dynasties. Her very effective attitude is not assumed by humans nor by other deities. Her interceding figure came into the religious art of Sumer through the influence of the liturgies in which the mother goddess is invariably represented as

1 For the date of the end of the Ur dynasty, see Thureau-Dangin, Rev. d'Assyr., xv, 47.

wailing for the sorrows of humanity.1 Seals now are engravedalmost invariably with the figures facing the right and the right hand of the worshipper raised as described. This fanciful processional scene continues throughout the seals of the Ur period.2 From it we may perhaps conclude that the worshippers actually stood with right hand raised and the left folded at the waist. In other words this pose in prayer persists from prehistoric times. But for some unexplained reason a second attitude begins to be assumed about the age of Dungi (2456-2399), illustrated by our Fig. 12. The secondary position of pre-Sargonic times is again assumed. The man's conducting god or goddess disappears; only the liturgical figure of the mother goddess standing behind the penitent remains. This attitude is exceedingly popular in the school of Ur, and may be seen on a great number of seals.3 What can be the influence which caused this important change in the attitude of prayer ? In the preceding pages the position was explained as one of humility and penance. Its prominence here is probably to be attributed to the same religious movement which introduced the figure of the interceding mother goddess, namely, the influence of the great liturgical schools which emphasized the element of sorrow and penance. These lugubrious litui'gies seem to have held powerful sway over mankind in that period. Archaeological evidence undisputably confirms two

1 See the writer's Babylonian Liturgies, xl, and especially Tammuz and Ishtar, p. 111. In the Cassite and later period (pre-Neo-Babylonian) the engraved kudnrrn frequently represent the seated goddess Gula with hands raised in the same way. 2 A bas-relief of a processional scene occurs on the stone tablet of Nabuapaliddin, king of Babylon 890-854 u.c., published in V Raw. GO. The relief is apparently modelled after a very ancient one, probably of the age of Ammizaduga, see Vordera$iati.iche Bibliothek, vol. iv, p. 50, by the writer. 3 See Delaporte, Bibliotheque Rationale, Nos. 98-128. Collection de Gltrcq, 171. Muse'e Guimet, 42-4.

accepted attitudes of prayer down to the end of the Ur dynasty. Of these the more ancient and important is the raised right arm, forearm parallel to the face, and palm inward. Left arm folded at the waist. This is the pose exclusively adopted henceforth in Babylonia. It is the attitude assumed by the great Hammurabi on the famous stele of Susa. In the majority of cases the palm is certainly turned inward.1 A tendency to turn the forearm and wrist so as to bring the palm at right angles with the face manifests itself on a few seals.5 This is usually the position seen on seals of the Cassite period, as our Fig. 13 shows.3 At this point we may arrest the archaeological discussion and turn to its literary aspect. The orthodox pose in prayer in Babylonia was the one imposed by Sumerian practice. The person stood left foot forward, right arm extended at right angle, forearm raised parallel to the body, palm inward or side-wise, left arm folded at the waist. It may be defined as the Sumerian custom. This accords with the Sumerian word for prayer, su-il-la or su-il-la-kam, translated into Semitic by nis fcati, " the lifting of the hand." i No examples of Sumerian su-il-la prayers have survived from the classical period.5 We do not know precisely how the Sumerians employed

1 Delaporte, Bibliothique Nationals, 129-150. Collection, de Glercq, 169, 170, 172. 2 For example, Ward, Morgan Collection, No. 98 ; Delaporte, Bibliotheque Nationale, 151 (left hand raised due to archaic influence ; the engraver adopted the old method showing the raised hand farthest from the observer). It is the poise of the hand assumed by Hammurabi on the Susa ttile. 3 Note also the position of the hand of Belahe-erba, an official of Merodachballadin (end of eighth century), as he stands before his king on the well-known Icudurru of Berlin (Hinke, A New Boundary Stone, p. 72 ; see also p. 23). 4 See IV Raw. 20, 9 ; ASKT. 127, 57 ; IV R. 17a, 53. 6 IV Raw. 53,. iii, 44- iv, 28, a list of forty prayers of this kind. The list is restored by a text, K. 3276, published in the writer's Babylonian • Liturgies, No. 103.

these prayers, but evidence points to their original use as pure orisons and unconnected with magic ceremonies or incantations. The Sumerians, of course, possessed long incantation ceremonies, as we know from unpublished tablets in Constantinople. But they did not associate pure prayer with incantation as the Semites did. We possess one Sumerian hi-il-la or prayer of the lifting of the hand with a Semitic translation, and this is identical with what we know to have been the penitential prayers er-sag-hun-ga, or "weeping to appease the heart". We know that the Sumerians employed a large number of the latter kind of prayers. Inasmuch as the texts repeatedly confuse the two, we may assume that both forms of prayer were confined to the private worship in distinction from the liturgies and er.semma hymns, or compositions for public service. The details concerning the Sumerian hymnal technology are intricate, and the discussion has been prolonged in the article on " Prayer (Babylonian)" in Hastings' Encyclopaedia of Religion and Ethics. My opinion is that the erktghunga prayers, commonly called "penitential prayers", were sung at a private service by a priest, whereas the hi-il-la prayers were recited by the persona orans. It is possible that the attitude with hands folded at the waist is the one assumed by the worshipper in the more formal ersaghunga. In any case the common attitude supplied by archaeological remains is clearly the one which gave the name to the ordinary prayers of the " lifting of the hand ". It is riot possible to study the archaeology of seals after the Cassite period in Babylonia. An age of intei'nal disorder marks the history of Babylonia from the twelfth to the end of the sixth century. When Nabopolassar restored a stable kingdom in 625 B.C. Babylonian glyptic had become a servile copy of Assyrian sphragistic. Now we have almost exclusively the signet type of seal as it was used in Assyria. On few of these

late seals does the old Babylonian pose in prayer survive.1 Although the Assyrian religion borrowed its prayers exclusively from the Babylonians, and employed the same names for them in both Sumerian and Semitic, nevertheless they usually modified the kiss-throwing raised hand, and also retained the old Semitic open-hand pose. Be it first remarked that when the Assyrians employed the phrase "lifting of the hand" for prayer the term was purely technical, borrowed along with the prayers themselves, and frequently applied to another attitude of orisons. The Assyrians retained the spirit and cult practices of Semitic religion much more than their kinsmen of the south. The Sumerian poses in prayer are not the only ones here. By far the most common attitude of prayer in Assyria is illustrated on Figs. 14 and 15. The feet are in the same position as in Sumer, but the left arm, instead of being folded at the waist, is extended at right angle with the body from the elbow. The right arm is properly in the same position (see Fig. 14) as in Sumer, but the forearm and hand turn outward, the index finger pointing at the god or sacred object with thumb closed over the three remaining fingers.2 This attitude is similar to one seen in Greece, and is the kiss-throwing hand in a new position. The pointed hand is a late motif which I shall discuss below in connexion with a similar Greek pose.3 On Fig. 16 is seen what I believe to be the true Semitic attitude of prayer of the Assyrians and Hebrews. Here both hands are extended

1 The following Assyrian seals probably belong in realitj' to Babylonia, and are to be assigned to the period of Shamash-shum-uktn, Collection de Clercq, 373, 372. Ward, Seals of the Morgan Collection, 145, assigns a seal with the old Sumerian processional scene to Assyria ; it is difficult to understand why Ward came to this conclusion. If the seal really be early Assyrian we have an example of Sumerian influence in Assyria. 2 See also Delaporte, Bibliotheqtie Nationale, 327, 330 et passim. 3 I refer naturally to the figure of the human here and elsewhere. The pose of the deities does not concern this discussion.

palms inward in the act of receiving blessings from the deity.1 Unfortunately all early Assyrian seals are devoted to other symbolism, consequently we cannot arrange the types of religious pose in chronological order. Fig. 17 shows a seal of an Aramaean, on which he represents himself as a eunuch in the same position as on the previous seal. The inscription reads, "Belonging to Akdeban, son of Gebrod, the eunuch who worships Adad. "2 We have here undeniably pure Semitic influence and type. On the same seal to the right stands a priest (?) in the Assyrian attitude of prayer denned above. Naturally the more widely adopted position with left arm held to the side and forearm extended palm upward is due to influence of the open-hand position or true Semitic type. On Fig. 1, another Aramaean seal, the owner, Jarp'el, stands in an attitude which combines the two positions. The right hand has the true Semitic pose ; the position of the left hand is influenced by the move common Assyrian attitude. On a seal of the Musee Gtt,i7net3 the owner stands in the Assyrian pointed-finger attitude, but the eunuch has the Assyrian and Semitic open- hand position.4 Excellent examples of the open-hand pose will be seen on No. 327 of the Collection de Clercq, No. 113 of the Musee Guimet,*Ro. 312 of the Bihliotlieque Rationale.5 In the Neo-Babylonian and Persian periods on the signet type of seal the open-hand position again prevails to the exclusion of all others. Figs. 18 and 20 will

1 The Stele of Seiiecherib represents that king with pointed right hand in praying position. The left hand is in the usual Assj'rian position, elbow at left hip, but here it holds a short sword. See R.A. xi, 189. 2 See also Collection de Clercq, 326 bis. 8 No. 109 of Delaporte's Catalogue. 4 O. A. Cooke, North Semitic Epigraphy, pi. xi, 2. 6 Last two numbers refer to Delaporte's Catalogues.

illustrate this period. The consensus of evidence, therefore, indicates that the peculiar Assyrian attitude of prayer with pointed index finger was sporadic, only partially true Assyrian, and was abandoned in later times by the Babylonians, who adopted the open-hand position. This conclusion, which has been made on archaeological remains alone, is effectively corroborated by Assyrian texts. In their native vernacular the term for to pray was " to open the hands", not " to raise the hand" as in Sumer and Babylonia. This philological commentary upon the archaeological evidence is so important that I introduce it here in the body of this discussion. A hymn to Nebo by Asurbanipal has the following line:— ip-te-te ilu Asur-bani-apal up-ni-Su it-ta-na-ah-har a-na ilu NabU bSli-Su. " Asurbanipal opened his hands and presented himself before Nebo his lord." (Craig, R.T., i, 6, rev. 1.) In an Assyrian letter a priest writes to his king:— up-ni-ia ap-te-ti ildnini us-sa-ar-ri-ir. " I have opened my hands and prayed to the gods." (Harper, Letters, No. 23, rev. 6.) An Assyrian hymn to Ishtar has the following passage:— pa-ta-ni up-na-ia-a a-na ilat Be-lit same* u-sal-la. " My hands are opened and I pray to the queen of heaven." (K. 890, published in transcription only by Arthur Strong, B.A., ii, 634.) The same attitude was assumed by vanquished princes before their Asssyrian conquerors:— 'u-a-ai zar-biS i-balc-ki-ma pi-ta-a ub-na-a-hju u-sal-la-a be-lu-ti. "(He ascended the wall of his city) wailing bitterly 'oh woe !' and his hands were opened as he prayed to my lordship." (Inscription of Assarhaddon describing JRAS. 1919. 36

the capture of the king of Supria, Winckler, Forschungen, ii, 28.) The son of a defeated king of the land of Mannai prays to Asurbanipal as follows :— up-na-a-su ip-ta-a u-ml-la-a belu-u-ti. " He opened his hands and implored my lordship." (Rassam Cylinder, iii, 17, in V Raw., pi. iii.) The open-hand position was, therefore, the traditional Semitic attitude of prayer as illustrated by Assyrian, late Babylonian, and Aramaic seals. This was undoubtedly the attitude among the Hebrews. Here we have almost no archaeological evidence. The seal (Fig. 21) from Moab east of the Jordan is the only design of this kind hitherto discovered in Palestine. Here only the right hand is represented in the half-elevated open .pose. The inscription reads, "Belonging to Eliamas, son of Elisha'."1 The evidence of the Hebrew texts is, however, con- clusive. The expression here is paras eth-kappim, " he spread out the hands." It occurs in the earliest Hebrew sources as Exodus ix, 29 ; ix, 33 ; Isaiah i, 15 ; Job xi, 13, etc. The attitude is one naturally assumed by those who look upward to heaven. Thus we have its essentially original implication in the description of Solomon's adoration of Jahweh, 1 Kings i, 22, "And Solomon stood before the altar of Jahweh in the presence of all the congregation of Israel and he spread out his hands to heaven." Undoubtedly this Semitic position in prayer 1 A stele with Aramaic inscription found at Nerab, south-east of Aleppo, in Syria, has the bas-relief figure of a priest of the moon-god. The priest's name, Sin-zer-ibni, clearly reveals his Assyrian origin. The right hand is raised in the half-turned kissing-hand pose, the left hand being held in Assyrian fashion, elbow at the hip, forearm straight forward, holding a wide double-edged knife. The date is probably post-Assyrian and of the period of Nebuchadnezzar and Astyages. Clearly the figure is under Assyrian and Persian influence, and cannot be used in the discussion of West Semitic religious gesture. For the inscription see Cooke, North Semitic Inscriptions, 186, and for the figure, Clermont-Ganneau, Album d'Antiquitts Orientates, pi. i.

emphasizes the reception of grace and favour from a deity who dwells in the sky. That can easily be accounted for in Hebrew. They regarded their god, Jahweh, as a sky- god from the beginning. The raised open-hand position is assumed when praying unto a far-away celestial deity ; prayer before a statue naturally demands a different pose. Since in pure Hebrew worship statues of the national god were forbidden we consequently find this to be the orthodox position among them. Inasmuch as the Hebrew pose in orisons appears also to have been the original Assyrian custom, we may assume that it was common to Semitic peoples. In Assyria the position is assumed before statues and divine symbols which is explained by the intrusion of the open-hand pose toward far-away gods into the cults as practised before statues. We have already observed that the Babylonians adopted this Assyrian position from the sixth century onward. It will be remembered that here the palms are held upward. Emphasis is here laid upon the desire to receive gifts from the deity. The Semitic open-hand pose is the one also employed by the Greeks and Romans when supplicating sky-gods. It is the so-called supines manus of the Romans; ccelo supinas si tuleris manus, " If thou liftest unto heaven thy hands bent backward," says Horace to the rustic maid Phidyle (Odes, Bk. iii, 23, 1). The same attitude was prevalent in Greek religion, where a number of expressions for lifting the hands (avaieiveiv ras ^etpa?) are common from Homer onward. Here, as in Assyria, the open-hand attitude may be assumed before statues, and the idea may be expressed by holding out only one hand in this attitude, derived from the custom of employing the other hand to present a cup of libation to the statue or sacred object. We have noted the same one-hand pose on the seal from Moab. We may assume that the Semites, the Greeks, and

Romans all worshipped in the open-hand position, arms raised, palms bent inwards. The position expresses supplication. This is the pose assumed by the Greek verb i/cereveiv, to approach (i/cco) in prayer, to supplicate. The psychology of this attitude is entreaty, and in this discussion it will be designated as the open-hand pose, or gestus supplicationis. Now we have seen that the Babylonians, although the oldest and most important branch of early Semitic races, did not employ the supplicatio position in prayer until the Neo-Babylonian period, when Assyrian influence prevailed. That is due to their complete adoption of Sumerian ideas and customs. The expression of religious emotions by means of the hands is profoundly different in Sumerian. Here the kissing hand so prevalent in Greece and Rome prevails from the very earliest period. The act of throwing a kiss to a statue or sacred object is equally primitive in Sumer. In Greece the gesture gave a word for the act, irpo^Kweiv, " to throw a kiss," to adore, to venerate. The Latins render this idea by adorare, venerari. It is the second great hand movement in religious psychology, and fundamentally conveys the idea of salutation, greeting, adoration. I shall designate this as the " kiss hand" pose or gestus adorationis. In our previous analysis of the Sumerian archaeology we saw that the processional scene is probably the oldest Sumerian idea of pose in worship. If a priest assumed the r61e of the conducting deity, then the design is not fanciful but real. At any rate the processional scene, which is based on the idea of a divine intercessor, occurs only in Sumerian and Egyptian religious art, and is another convincing strand of relationship between these two great religions. If the reader will refer to Fig. 2 he will see three characteristically Sumerian attitudes of worship. The seal is prehistoric and may be dated at

about 3500 B.C. The second figure from the right carries in his left arm a kid, and with his right arm he throws a kiss. The hand is here placed in the position of just having left the lips, and is retained in this position throughout Sumerian art. The first figure on the right has the hands folded at the waist. That is the second great attitude of hand expression, and denotes humility, submission, contrition, the fundamental and characteristic sentiment of Sumerian religion.1 It is apparently unknown in other ancient religions of the Mediterranean area, except in so far as it was adopted by the Babylonians and Hittites. We have seen that the folding of both hands at the waist was particularly common in the period of Dungi in the twenty-fifth and twenty-fourth centuries B.C., after which period it completely disappears in Babylonia. The fact that this pose continues in Hittite religion can lead to only one conclusion, namely, that the Hittites came under Sumerian influence in Asia Minor and Anatolia in the twenty-fifth century B.C., when the Sumerian dynasty of Ur is known to have been recognized in this region.2 The exact method of closing one hand upon the other cannot be detected on seals, for the execution is too minute. But it can be studied in detail on the statues of Gudea.3 The right hand is clasped by the left hand in an extraordinary manner, so that the right thumb lies against the body and right fingers lie almost horizontal. The position is physically impossible, and only an exaggeration of the natural clasp seen in the bas-relief of Asarhaddon at Sendschirli.4 It is curious

1 See the writer's Babylonian Liturgies, pp. xxxix f. 2 Thureau-Pangin, Rev. d'Assyr., 8,144, contract sealed by a cylinder dedicated to Ibi-Sin. Note also the Cappadocian seals on plate i which accompanies Th.D.'s article; here the poses in prayer are those of the period of Dungi. 3 See, for example, Heuzey, Antiquities Ckaldiennes, No. 50. 4 AusgrabungeninSendschirli,T$.onig\iche~Museenzu Berlin,Mitteilungen aus den Orientalischen Sammlungen, Heft xi, Tafel iii.

that the clasped-hand position should have survived only in bas-reliefs in Assyria.1 In early seals the processional scene represents the human, who is led by a deity, with the disengaged arm folded at the waist. See also Fig. 4. In later times, from Gudea to Dungi, the disengaged hand is in the kissing position (Fig. 11). This of course means that in early Sumerian religion the conducted worshipper approached with folded hands, whereas in later times he approaches in the act of adoration. We have noted how the folded-hand position obtained great popularity in the period of Dungi (Fig. 12), and finally disappeared altogether in favour of the kissing-hand position with one arm folded at the waist. This widely adopted attitude of Babylonian religion seems to have been introduced by the Semites of the first dynasty as a simple means of com- bining the two principal religious poses of the Sumerians. They thus combined the ideas of salutation and humility. We have also noted how in Babylonia down through the Cassite period there is a tendency to arrest the kiss- throwing hand in a position of half turn towards the god's statue. The gesture was continued in Assyria and in Western Semitic centres under Assyrian influence. We are now in a position to understand the extraordinary pointed-finger attitude of the Assyrians (see Fig. 14), as they worshipped before statues and sacred symbols. It is really the kiss-throwing hand arrested in the last stage of the act and thrown with the index finger only. The thumb is closed over the other three fingers, and the index finger lies on the thumb.2 The left hand instead of being retained in the folded position remains in the open-hand position common to the Semites. Here also we have a combination, but a combination of

1 See the Louvre bas-relief of the palace of Sargon, Assyrian Sculptures, Kleinemann & Co., plate vi, grand vizier and eunuch. 2 This Assyrian position can best be seen in Assyrian Sculptures, part i, photograph on the cover.

a Sumerian or Greek pose with a Semitic pose, the whole denoting salutation and supplication. Now this Assyrian transformation of the old Sumerian kissing hand with palm inward and fingers bent near to the lips has a parallel in Greece. Here also the kissing hand begins in exactly the same position as among the ancient Sumerians.1 Then the index finger is put to the lips, the thumb being closed over the other three fingers as described above.2 But the Greeks also transformed the attitude, arresting it in the act of throwing the kiss, or the pointed-finger position.3 Other forms of the kissing hand in Greece occur. For example, the bas- relief of the Apotheosis of Homer, fourth register, represents the adorers with hand thrown outward or the original gesture arrested in the last stage.4 Occasionally the hand is held to the lips with fingers slightly drawn together to the thumb.5 The Assyrian attitude with pointed index finger is, therefore, prevalent in Greece, and may possibly be due to Assyrian influence in Greek lands. In any case the kiss-throwing hand seems to have been an attitude common to Semitic peoples. If we may assume the Assyrian pointed-finger pose as true to type we may conclude that this was the pose employed by the Ba'al worshippers of Canaan or the early Semitic inhabitants

1 See Baumeister's Denkmciler des Klassischen Altertnms, 592. 2 This is described by Appuleius in his Metamorphoseon (iv, 28), as follows: "admoventes oribus suis dexteram primore digito in erectum pollieem residente" (Placing the right hand to their lips, the index finger lying upon the erect thumb). Professor Percy Gardner is inclined to regard this index-finger attitude in Greece as Oriental, and Appuleius is known to have been under Asiatic influence. But see the bas-relief of the Apotheosis of Homer (Tafel iv of Carl Sittl's Die Gebarden der Griechen und Romer), a good classical work of Greek art; in the fourth register the figure of Mnene has the position described by Appuleius. 3 Baumeister, ibid., 297, fig. 312. Under Oriental influence? 4 Baumeister, ibid., 112. 5 Overbeck, Gallerie heroischer BildwerJce, xxviii, 7. For the pointed finger see Overbeck, ibid., xi, 8.

of Palestine whose pagan worship so often contaminated the monotheistic religion of their conquerors, the Hebrews. The classic reference to this Canaanitish custom is found in the story of the anointing of Jehu to be king over Israel by Elijah. The narrative in 1 Kings xix, 18 describes the idolatry which had pervaded the Hebrew worship. "And I have left in Israel seven thousand, all the knees which have not bowed to Ba'al and every lip which has not kissed him." These events occurred in the ninth century B.C. From this passage alone one might infer that the Canaanites actually kissed the statues of Ba'al, but a passage in Job, which refers to the same worship several centuries later, admits no doubt concerning the nature of the act. " And if I beheld the luminary how it glows, and the moon passing in splendour, and my heart became perverted in secret, and my hand kissed my lip."x The kiss-throwing hand, therefore, persisted in the old pagan religion of Canaan probably as long as Ba'al worship survived. It was rejected by the Hebrews as wicked and sinful because it was connected with the worship of images. In Egyptian religion the orthodox attitude is entirely different. Here both hands are held out and upward, palms outward. It is the pose described by the Romans as manibus passis implorare, " to implore with out- stretched hands."2 The kissing hand in Egyptian religion is doubtful, but in the processional attitude the worshipper who is led by a deity holds his disengaged arm in a position which does suggest the kiss-throwing hand.8 We have no means of tracing religious poses in

1 Job xxxii, 26 f. a For examples in Egyptian religion seethe Book of the Dead, Papyrus Ani, 2nd ed., facsimile by Budge, pi. ii tt passim. For the pose in Roman religion see the Louvre statue in Bouillon, ii, 29. 3 See ibid. pi. iv, the hawk-headed Horus leads a worshipper to Isis. Egyptologists whom I have consulted unanimously regard the kissing hand as unknown in that religion.

prayer in the early stages of Egyptian religion. All the known representations are from the eighteenth dynasty and later.1 Since the kiss-throwing hand was obviously the most important gesture of worship in Sumerian, Babylonian, and Assyrian religion we must surely expect to find the idea reflected in the language itself. We have seen that the Sumerians described the gesture by the colourless phrase su-il-la, "lifting of the hand."2 Since only one hand is referred to in this term, the description clearly refers to the orthodox position of the right hand as assumed either in processional scenes before the first dynasty or in the Hammurabi pose which supplanted all others. But the original meaning of the gesture is not reflected by this phrase, possibly because it had been forgotten. We are, however, able to trace the attitude both in early Sumerian pictographs and in Sumerian philology. The two sources of interpretation are closely connected, and are discussed together here.

\\\\ Ideogram for "to salute with a kiss, to adore". Sumerian sub, Sub = naSdku, to kiss, and ikribu, prayer.

A pictograph for prayer occurs on the earliest monu- ments, and consists of the sign for mouth or face with

1 But note the hieroglyphic determinative for verbs of praying in

Egyptian V . This hieroglyph is extremely ancient. Since it

means " to pray", the hieroglyph is obviously based upon the orthodox gesture in prayer. (Note by Dr. Blackman.) 2 Translated into Semitic by ni$ kali.

the sign for hand inserted into it.1 This ideograph has the Sumerian value sub, sub, the word for " to kiss ",2 which already in the earliest period came to mean " pray, prayer".3 Since the kiss-throwing hand is the earliest and most important prayer gesture, this ideograph obviously refers to that attitude. A statue of the famous Rim-Sin, king of Larsa, was set up in the temple of the sun-god, and represented him saying his prayers in the kissing-hand pose,4 precisely as we see Hammurabi on the Susa stele. The same description is given for a statue of Samsuiluna, king of Babylon in the same period.5 In liturgical worship of the public services prostration before the statues of the deities must have been ritualistically imposed, at least for the psalmists. One of the rubrics of liturgies is ki-isub, which means literally "to kiss the earth", and is translated into Semitic by sukSna,6 to bow down, or by kakkara nussuku, /' to kiss the earth."7 This rubric occurred originally after each song of the service, and clearly indicates that at some point in the singing a prostration, or at least a deep bow, was made to the deity.8 The ideograph and word sub, sub discussed above commonly mean " prayer ", and may be employed as a general word for prayer, but the original and strictly proper sense of both is to salute by the hand kiss. Another Sumerian expression for a gesture in prayer is 1 The sign KA + hi will be found in Thureau-Dangin, REG No. 198, and its early forms and meanings discussed in PSBA. 1911, 50-2. 2 In all known texts the verb is written su-ub = naSdku, IV Raw. 9a, .-.9 ; K. 5098, obv. 4 ; PSBA. 1911, 88, 40. 3 sub means both Icardbu and ihribu. 4 Salan urudu sub-sub-be d-Ri-im-d-Sin, " Two copper statues of Rim-Sin in praying (kissing hand) attitude," RA. 15, 7, 12. 5 PBS. x, 152. 6 See Geller, ATU. i, 306, 11, kia-ge-su-ub = hi tmklna. Finally, sub-sub came to mean "bow down ". 7 IV Raw. 9a, 59. 8 Later kisub, ki-Hu-bi-im came to mean "strophe" simply. See BL. p. xlv.

commonly supposed to mean " prostration of the face to the earth". It has been so interpreted because the Semites invariably translate the expression by laban appit "to prostrate the face." The usual Sumerian term is ka-Su-gdl, and occasionally lca-su-tag} Neither of these terms conveys the idea of prostration. The second term clearly means " to touch the mouth with the hand". The first term is more difficult, for no suitable meaning can be found for the verb gal} Since, however, gals means nasu, "to lift," the term may possibly mean "to lift the hand to the mouth ". The original meaning of this phrase probably refers to the kiss-throwing hand also. At any rate the Sumerians clearly attached some special meaning to this act. In classical texts it invariably describes prayer in a temple before a statue.4 It also describes the attitude of a penitent when he recites the penitential prayers called ersaghunga.5 In my previous pages I suggested that the folded hands may be the pose assumed here. In any case this term, ka-su-gal, although originally referring to the kissing hand, clearly denotes bowing, at least in actual usage. It is employed of reducing foreign lands to submit to the yoke of

1 lea Su-ga-ra-ab-tag-gi-ne = appa-Sina lilbinalcum, "maythey bow their faces to thee," King, LIH. iii, 174,16. Of Innini the king Ishme-Dagan says, ha isu-ga-ra-ah-tag-yi, "I will bow the face to thee," PSBA. 1918, 54, 15. 2 The meaning labdnu for gal is derived by Semites from their own rendering of ica-Su-gdl. 3 The variant ma-al occurs, and since ma-al visually stands for gdl = isalcdnu, to institute, cause to be, perhaps the meaning of the phrase is rather "to place the hand to the mouth". Either of these interpretations is difficult since they violate Sumerian syntax, the object coming after the locative noun. The choice of interpretation given in the note is strengthened by the form Tca-Su-mar-ra-mu — laban appi-ia, IV R. 20, 9. * SAK. 42a, v, 4 ; Gudea, Cyl. A. 18, 9 ; 8, 14 ; in B. 8, 19 ka-mi-gdl is employed for the intercessory prayer of a god to the great deity Ningirsu on behalf of the patesi Gudea. The term is employed in the same sense in Clay, Miscel., p. 6, iv, 9. s IV Raw. 26, 62, and Var. 27, 36.

a conqueror, and could hardly have obtained that sense unless the act consisted in some such gesture.1 On the other hand it is employed for the act of an interceding god who conducts his protege by the hand, and cannot possibly refer to prostration.2 Senecherib, king of Assyria, relates in his Bavian inscription how he made a bas-relief of himself on the rocks of the Tas mountain; and this image of himself stood in the attitude (Idbin appi) commonly supposed to indicate prostration. But the sculptured rock of Bavian represents Senecherib standing upright in two attitudes ; one, in which the figure is marred by a hole in the rock, represents him standing with arms folded at the waist.3 In any case the Semitic expression means simply supplication, prayer, and not prostration here.4 The term therefore became colourless, and meant simply supplication. It is obvious, however, that bowing,5 and even prostration,6 were actually added to the old gesture of salutation with the kissing hand. We should be cautious about deducing principles from the art alone. The sculptors and seal engravers of Sumer, Babylonia, and Assyria avoid the more humble attitudes by tradition, since the laws of ancient art are based upon primitive ideas and remain true to type. Kneeling is also mentioned in Sumerian and Babylonian

1 kur-ri ha-hi-gdl, "he that causes the foreign land to submit," PBS. x, No. 9, i, 7; ibid. 141, 4. 2 Gudea, Cyl. B. 8, 19. 3 See Hommel, Geschichte Babyloniens und Assyriens, 687. For th& inscription see Meissner-Rost, Bauinschriften Sanheribs 76, 55. Perhaps Senecherib's inscription at Bavian refers to the scene on the rocks which have fallen into the river, a drawing of which is given in Layard's Nineveh and Babylon, p. 72. Here the king stands before Asur and has the ordinary kiss-throwing hand pose. 4 W. Schrank, Babylonische Siihnriten, p. 58, takes the old view and regards prostration as the original liturgical sense of ha-Su-gdl and labdn appi. 5 Sulcenu. 6 labdn appi.

religious texts, but it does not appear, to have been employed in the liturgical worship. As prostration was one of the acts of the penitential psalms (ersaghunga), so kneeling appears generally to have been employed in the private prayers of the lifting of the hand (su-il-la). This latter class of prayer was much more informal, and in Babylonian religion usually formed part of the magic rituals of private atonement. " I kneel, I stand, seeking thee," says a prayer to the moon-god employed in the ritual of the house of baptism.1 Another text prescribes as follows : " On his knees thou shalt cause him to bend, and at the right side of a copper statue thus he shall recite."2 The text is a ritual for the purification of a house. The word 3 here translated " knee" really means the tibia, hence the attitude is unmistakably identical with the position assumed in Christian prayer. A prayer of Asurbanipal is introduced by the rubric: " Kneeling upon his knees4 Asurbanipal presented himself before Nebo his lord." In a ritual for the purification of a king the rubrics order that the king shall kneel5 when he recites various prayers of the lifting of the hand. Asurbanipal describes his preparation for an improvised prayer to Ishtar for help in battle : " I stood over against her and kneeled at her feet."6 The attitude is mentioned once in a Sumerian penitential psalm, and the gesture may possibly have been accepted in this more formal kind of prayer also.7 The Semites invariably employed Sumerian in this class of prayer, but provided their texts with a Semitic translation. 1 King, Magic, i, 21, kamsaku azzaz aSe'ka kaSu. 2 Craig, RT. 66, 18. 8 kinm. 4 kamis ina kinsi-Su, Craig, RT. 5, 19. 6 ikammis-ma kiam ikabbi, IV Raw. 546, 20, 22, 24, 26, 28, 30, 42, 44, 46, 48 (here gam = ikammis). 6 akmes Sapal-Sa, Streck, Asurbanipal, ii, 112, 28. 7 i-di-zu mu-un-gam-dm — mahar-ka kansaku, psalm to Nergal or Ninurta, IV Raw. 246, 10.

Seals of the Cassite period actually represent a worshipper in the kneeling position, the hands being retained in the ordinary kissing position.1 One remarkable seal has the owner in two positions before his god, one in the ordinary standing pose and one in the kneeling posture.2 The artist draws the kneeling figures much smaller than the standing figures, thus showing that the posture conveyed greater humility. It seems on the whole certain that kneeling was regarded as an inferior and awkward posture by the Sumerians, and is a Semitic innovation in Babylonia. It is, as we have seen, uncertain whether the Sumerians even permitted prostration in formal worship. The only evidence consists in a phrase (ka-su-gal, ka-su-tag) which the Semites, probably under influence of their own custom, translated by "to prostrate the face", and even this translation refers to the kissing-hand gesture on the inscriptions of Bavian. In Assyria and Babylonia the attitude was largely confined to the informal prayers of magic rituals 3 and to impromptu prayer.4 On the whole we may say that the Sumerians held kneeling and prostration in much the same low estimate as did the Greeks who regarded the gesture as unworthy of freemen. It has been noted that the Babylonians when kneeling at prayer retained the ordinary gesture of the hand. The Hebrews follow the same custom when they kneel. Solomon kneeled before the altar and spread out his hands to heaven,5 thus retaining the orthodox Hebrew hand gesture. In the time of Elijah the Hebrews are

r Collection de Clercq, 264. 2 Ibid. 2S8. The half-kneeling figure on a seal of the Cassite period, Babylonica, iii, 238, is unique. 3 IV R. 60ft, 19, ashur-kunuSi ase'-kunusi Sapal-kun akmis, "I have turned to you, I have sought you, I have kneeled at your feet." 4 On kneeling, see also Schrank, Babylonische Siihnriten, 59 ff., and the writer in Babyloniaca, iii, 236 f. * 1 Kings viii, 54 ; 2 Chron. vi, 13

accused by Jahweh of bowing the knee to the Ba'als of the old pagan religion,1 which proves that the indigenous Semites of Palestine also employed kneeling in worship.2 Also Ezra kneeled upon his knees and opened his hands unto Jahweh in prayer,3 and the pose is spoken of by the writer of the Exilic Psalm 95 in a way which presupposes it as the orthodox attitude. Kneeling must therefore be regarded as essentially Semitic, and the Christian practice is certainly to be traced to that source. In India the clasped hands held palms facing each other, fingers outward, closed and pointed slightly upward, seems to be the only orthodox attitude of prayer in a formal sense. The standing attitude is maintained, the feet being held on the same plane, not in the attitude of advancing. For permission to reproduce drawings and photographs of seals the writer gratefully mentions the following publishers :— 1. Ernest Leroux, editor of Louis Delaporte's Catalogues of the seals in the Bibliotheque Nationale and the Musee Guimet. 2. The Clarendon Press, publishers of G. A. Cooke's North Semitic Inscriptions. 3. Maisonneuve Freres et Ch. Leclerc, editors of J. Menant's Recherches sur la Glyptique Orientate. 4. The Carnegie Institution of Washington, publishers of W. H. Ward's The Seal Cylinders of Western Asia. 5. The J. Pierpont Morgan Library, publisher of W. H. Ward's Catalogue of seals in that private collection. 6. J. C. B. Mohr, publisher of H. Gressmaim's Texte und Bilder.

1 1 Kings xix, 18. 2 That is, at any rate, a safe inference. 3 Ezra ix, 5. The Hebrew expression is precisely parallel to the Assyrian, but the words are different, ekre'd 'al-birkai, and Assyrian alcjnis ina kinse-d.

INDEX TO THE FIGURES ON THE ACCOMPANYING PLATE. FIG. 1. G. A. Cooke, North Semitic Inscriptions, pi. xi 3. 2. L. Delaporte, Catalogue, No. 62. 3. W. H. Ward, Seal Cylinders, No. 271. 4. W. H. Ward, No. 270a. 5. W. H. Ward, No. 276. 6. J. Menant, Glyptique, vol. i, pi. iii 4. 7. L. Delaporte, No. 81. 8. L. Delaporte, No. 79. 9. Revue d' Assyriologie, v, 135. 10. W. H. Ward, Morgan Collection, 52. 11. J. Menant, Qlyptique, vol. i, pi. iv 2. 12. L. Delaporte, No. 111. 13. L. Delaporte, No. 300. 14. L. Delaporte, No. 355. 15. L. Delaporte, No. 354. 16. W. H. Ward, Seal Cylinders, 686. 17. G. A. Cooke, North Semitic Inscriptions, pi. xi 2. 18. L. Delaporte, No. 532. 19. L. Delaporte, Musie Guimet, No. 145. 20. H. Gressraann, Altorientalische Texte und Bilder, Abbildung, 200. 21. = Fig. 20, enlarged.